Magnesium Inhibits Spontaneous and Iron-induced Aggregation of α-Synuclein
2002; Elsevier BV; Volume: 277; Issue: 18 Linguagem: Inglês
10.1074/jbc.m107866200
ISSN1083-351X
AutoresNatalie Golts, Heather M. Snyder, Mark Frasier, Catherine Theisler, Peter S. Choi, Benjamin Wolozin,
Tópico(s)Restless Legs Syndrome Research
ResumoMultiple studies implicate metals in the pathophysiology of neurodegenerative diseases. Disturbances in brain iron metabolism are linked with synucleinopathies. For example, in Parkinson’s disease, iron levels are increased and magnesium levels are reduced in the brains of patients. To understand how changes in iron and magnesium might affect the pathophysiology of Parkinson’s disease, we investigated binding of iron to α-synuclein, which accumulates in Lewy bodies. Using fluorescence of the four tyrosines in α-synuclein as indicators of metal-related conformational changes in α-synuclein, we show that iron and magnesium both interact with α-synuclein. α-Synuclein exhibits fluorescence peaks at 310 and 375 nm. Iron lowers both fluorescence peaks, while magnesium increases the fluorescence peak only at 375 nm, which suggests that magnesium affects the conformation of α-synuclein differently than iron. Consistent with this hypothesis, we also observe that magnesium inhibits α-synuclein aggregation, measured by immunoblot, cellulose acetate filtration, or thioflavine-T fluorescence. In each of these studies, iron increases α-synuclein aggregation, while magnesium at concentrations >0.75 mm inhibits the aggregation of α-synuclein induced either spontaneously or by incubation with iron. These data suggest that the conformation of α-synuclein can be modulated by metals, with iron promoting aggregation and magnesium inhibiting aggregation. Multiple studies implicate metals in the pathophysiology of neurodegenerative diseases. Disturbances in brain iron metabolism are linked with synucleinopathies. For example, in Parkinson’s disease, iron levels are increased and magnesium levels are reduced in the brains of patients. To understand how changes in iron and magnesium might affect the pathophysiology of Parkinson’s disease, we investigated binding of iron to α-synuclein, which accumulates in Lewy bodies. Using fluorescence of the four tyrosines in α-synuclein as indicators of metal-related conformational changes in α-synuclein, we show that iron and magnesium both interact with α-synuclein. α-Synuclein exhibits fluorescence peaks at 310 and 375 nm. Iron lowers both fluorescence peaks, while magnesium increases the fluorescence peak only at 375 nm, which suggests that magnesium affects the conformation of α-synuclein differently than iron. Consistent with this hypothesis, we also observe that magnesium inhibits α-synuclein aggregation, measured by immunoblot, cellulose acetate filtration, or thioflavine-T fluorescence. In each of these studies, iron increases α-synuclein aggregation, while magnesium at concentrations >0.75 mm inhibits the aggregation of α-synuclein induced either spontaneously or by incubation with iron. These data suggest that the conformation of α-synuclein can be modulated by metals, with iron promoting aggregation and magnesium inhibiting aggregation. Parkinson's disease Parkinson’s disease (PD)1Hoehn M.M. Yahr M.D. Neurology. 1998; 50: 318-334Crossref PubMed Google Scholar is a common motor disorder that affects about 1% of population over the age of 65 (1). The disease is characterized by progressive neurodegeneration predominantly affecting dopaminergic neurons in the nigrostriatal system (2Gibb W. Lees A. J. Neurol. Neurosurg. Psych. 1988; 51: 745-752Crossref PubMed Scopus (2840) Google Scholar). The degenerating neurons develop intracellular inclusions, termed Lewy bodies, which are composed of a dense core of filamentous and granular material (3Spillantini M.G. Crowther R.A. Jakes R. Hasegawa M. Goedert M. Proc. Natl. Acad. Sci. U. S. A. 1998; 95: 6469-6473Crossref PubMed Scopus (2443) Google Scholar). Recent studies indicate that α-synuclein is a major filamentous component of Lewy bodies (3Spillantini M.G. Crowther R.A. Jakes R. Hasegawa M. Goedert M. Proc. Natl. Acad. Sci. U. S. A. 1998; 95: 6469-6473Crossref PubMed Scopus (2443) Google Scholar, 4Spillantini M. VM-Y M. Schmidt L. Trojanowski J. Jakes R. Goedert M. Nature. 1997; 388: 839-840Crossref PubMed Scopus (6265) Google Scholar). Genetic studies suggest that α-synuclein plays a key role in the pathophysiology of PD, because mutations in α-synuclein, at A53T or A30P, are associated with early-onset familial PD (5Polymeropoulos M.H. Lavedan C. Leroy E. Ide S.E. Dehejia A. Dutra A. Pike B. Root H. Rubenstein J. Boyer R. Stenroos E.S. Chandrasekharappa S. Athanassiadou A. Papapetropoulos T. Johnson W.G. Lazzarini A.M. Di Duvoisin R.C. Iorio G. Golbe L.I. Nussbaum R.L. Science. 1997; 276: 2045-2047Crossref PubMed Scopus (6732) Google Scholar, 6Kruger R. Kuhn W. Muller T. Woitalla D. Graeber M. Kosel S. Przuntek H. Epplen J. Schols L. Riess O. Nat. Genet. 1998; 18: 106-108Crossref PubMed Scopus (3344) Google Scholar). The accumulation of aggregated protein underlies the pathophysiology of many neurodegenerative disorders, and increasing evidence suggests that aggregated α-synuclein plays a key role in the pathophysiology of PD. α-Synuclein has a strong tendency to aggregate and does so spontaneously in vitro at a slow rate (7Hashimoto M. Hsu L. Sisk A. Xia Y. Takeda A. Sundsmo M. Masliah E. Brain Res. 1998; 799: 301-306Crossref PubMed Scopus (249) Google Scholar, 8Wood S. Wypych J. Steavenson S. Louis J. Citron M. Biere A. J. Biol. Chem. 1999; 274: 19509-19512Abstract Full Text Full Text PDF PubMed Scopus (612) Google Scholar, 9Conway K.A. Lee S.J. Rochet J.C. Ding T.T. Williamson R.E. Lansbury Jr., P.T. Proc. Natl. Acad. Sci. U. S. A. 2000; 97: 571-576Crossref PubMed Scopus (1347) Google Scholar). Both the A53T and the A30P mutations in PD increase the tendency of α-synuclein to aggregate. Many studies in cultured neurons, and some studies in transgenic animals, suggest that α-synuclein aggregation is linked to cellular toxicity and neurodegeneration (10Ostrerova-Golts N. Petrucelli L. Hardy J. Lee J. Farrer M. Wolozin B. J. Neurosci. 2000; 20: 6048-6054Crossref PubMed Google Scholar, 11Masliah E. Rockenstein E. Veinbergs I. Mallory M. Hashimoto M. Takeda A. Sagara Y. Sisk A. Mucke L. Science. 2000; 287: 1265-1269Crossref PubMed Scopus (1572) Google Scholar, 12Feany M.B. Bender W.W. Nature. 2000; 404: 394-398Crossref PubMed Scopus (1725) Google Scholar). In cell culture, formation of α-synuclein aggregates correlates with cell injury (10Ostrerova-Golts N. Petrucelli L. Hardy J. Lee J. Farrer M. Wolozin B. J. Neurosci. 2000; 20: 6048-6054Crossref PubMed Google Scholar). Overexpressing α-synuclein in Drosophila leads to an age-dependent accumulation of aggregated α-synuclein and associated neurodegeneration (12Feany M.B. Bender W.W. Nature. 2000; 404: 394-398Crossref PubMed Scopus (1725) Google Scholar). Masliah and colleagues also observed that aggregated α-synuclein is associated with loss of markers in dopaminergic neurons, although other studies of α-synuclein overexpression in transgenic mice have been less conclusive (11Masliah E. Rockenstein E. Veinbergs I. Mallory M. Hashimoto M. Takeda A. Sagara Y. Sisk A. Mucke L. Science. 2000; 287: 1265-1269Crossref PubMed Scopus (1572) Google Scholar, 13Kahle P.J. Neumann M. Ozmen L. Muller V. Jacobsen H. Schindzielorz A. Okochi M. Leimer U. van Der Putten H. Probst A. Kremmer E. Kretzschmar H.A. Haass C. J. Neurosci. 2000; 20: 6365-6373Crossref PubMed Google Scholar, 14van der Putten H. Wiederhold K.H. Probst A. Barbieri S. Mistl C. Danner S. Kauffmann S. Hofele K. Spooren W.P. Ruegg M.A. Lin S. Caroni P. Sommer B. Tolnay M. Bilbe G. J. Neurosci. 2000; 20: 6021-6029Crossref PubMed Google Scholar). Thus, increasing lines of evidence suggest that aggregation of α-synuclein is associated with the degeneration of dopaminergic neurons and suggest that α-synuclein contributes to the neurodegenerative processes occurring in PD. Recombinant α-synuclein aggregates spontaneously following prolonged incubation in vitro. Recently, we and others have shown that α-synuclein also aggregates rapidly following exposure to Fe(II) (10Ostrerova-Golts N. Petrucelli L. Hardy J. Lee J. Farrer M. Wolozin B. J. Neurosci. 2000; 20: 6048-6054Crossref PubMed Google Scholar,15Paik S.R. Shin H.J. Lee J.H. Arch. Biochem. Biophys. 2000; 378: 269-277Crossref PubMed Scopus (173) Google Scholar). In vitro, Fe(II) accelerates the rate of α-synuclein aggregation. For example, similar amounts of aggregation are inducedin vitro by incubating 23 μm α-synuclein alone for 30 days or with 50 μm FeCl2 for only 3 days, suggesting that 50 μm Fe(II) increases the rate of α-synuclein aggregation about 10-fold (see discussion below). These observations suggest that interaction with iron could greatly accelerate α-synuclein aggregation. The factors regulating α-synuclein aggregation in the brain are poorly understood. Some studies suggest that neurotoxins, such as the pesticide rotenone or paraquat, stimulate α-synuclein aggregation (16Li V.N. Uversky J. Fink A.L. FEBS Lett. 2001; 500: 105-108Crossref PubMed Scopus (43) Google Scholar). The involvement of metals in PD suggests that metals might also play a role in the aggregation of α-synuclein and the pathophysiology of PD. Epidemiological studies have shown that exposure to metals is associated with PD. For instance, individuals with industrial exposure to iron, copper, and/or lead have high rates of PD (17Gorell J.M. Johnson C.C. Rybicki B.A. Peterson E.L. Kortsha G.X. Brown G.G. Richardson R.J. Neurotoxicology. 1999; 20: 239-247PubMed Google Scholar). Neuropathological studies show that synucleinopathies are generally associated with iron accumulation, which is consistent with a pathological link between iron and α-synuclein (18Duda J.E. Lee V.M. Trojanowski J.Q. J. Neurosci. Res. 2000; 61: 121-127Crossref PubMed Scopus (250) Google Scholar). Brains from patients with PD, type I iron storage disease (Hallorvorden-Spatz disorder), and multiple systems atrophy all show increased iron content (19Galvin J.E. Giasson B. Hurtig H.I. Lee V.M. Trojanowski J.Q. Am. J. Pathol. 2000; 157: 361-368Abstract Full Text Full Text PDF PubMed Scopus (170) Google Scholar). In PD the levels of iron are increased over controls, and Fe(II) has been identified as a major component of Lewy bodies (20Castellani R. Siedlak S. Perry G. Smith M. Acta Neuropathol. 2000; 100: 111-121Crossref PubMed Scopus (202) Google Scholar, 21Mann V.M. Cooper J.M. Daniel S.E. Srai K. Jenner P. Marsden C.D. Schapira A.H. Ann. Neurol. 1994; 36: 876-881Crossref PubMed Scopus (216) Google Scholar, 22Griffiths P.D. Dobson B.R. Jones G.R. Clarke D.T. Brain. 1999; 122: 667-673Crossref PubMed Scopus (198) Google Scholar, 23Dexter D. Carayon A. Javoy-Agid F. Agid Y. Wells F. Daniel S. Lees A. Jenner P. Marsden C. Brain. 1991; 114: 1953-1975Crossref PubMed Scopus (871) Google Scholar, 24Barbiroli B. Martinelli P. Patuelli A. Lodi R. Iotti S. Cortelli P. Montagna P. Mov. Disord. 1999; 14: 430-435Crossref PubMed Scopus (51) Google Scholar, 25Uitti R.J. Rajput A.H. Rozdilsky B. Bickis M. Wollin T. Yuen W.K. Can. J. Neurol. Sci. 1989; 16: 310-314Crossref PubMed Scopus (111) Google Scholar, 26Durlach J. Bac P. Durlach V. Rayssiguier Y. Bara M. Guiet-Bara A. Magnes. Res. 1998; 11: 25-42PubMed Google Scholar). How iron contributes to Lewy body formation and the pathophysiology of PD, though, is not understood. In the experiments described below, we examine the regulation of α-synuclein aggregation using both spontaneous and iron-induced α-synuclein aggregation in vitro and show contrasting actions of iron and magnesium on α-synuclein aggregation. These studies have important implications for the pathophysiology of PD and other synucleionopathies. α-Synuclein (wild-type, A53T, and A30P) was cloned into the NcoI/NotI site of the Pro-Ex His 6 vector (Invitrogen). To generate recombinant α-synuclein, BPer (Pierce) reagent was used to solubilize the recombinant α-synuclein from the isopropyl-1-thio-β-d-galactopyranoside-induced bacterial lysates, which were then passed over a nickel-agarose column for purification. All spectrophotometric analysis were repeated three to five times. Cells were harvested with SDS lysis solution (2% SDS, 10 mm Tris, pH 7.4, 2 mm β-glycerol phosphate, 1 μm AEBSF). The amount of protein was determined using the BCA assay (Pierce), 5–30 μg per lane was run on 14% SDS-polyacrylamide gels and transferred to nitrocellulose (200 mA, 12 h). The nitrocellulose was then incubated 1 h in 5% I-block (Tropix)/phosphate-buffered saline, washed, incubated overnight in primary antibody, washed, then incubated 3 h in peroxidase-coupled secondary antibody and developed with chemiluminescent reagent (PerkinElmer Life Sciences). For analysis of aggregation using thioflavine-T, 23 nm α-synuclein was incubated in 10 μm thioflavine-T (in 50 mm glycine, pH 8.5) and measured by fluorescence (λex = 440, λem = 450–600 nm). To analyze aggregation of α-synuclein by filtration, samples were diluted into 100 μl of water, filtered through cellulose acetate (0.2 μm pore size), washed with 200 μl phosphate-buffered saline, and then immunoblotted as described above. To understand factors regulating α-synuclein aggregation, we investigated the interaction of different metals with α-synuclein using tyrosine fluorescence (27Garzon-Rodriguez W. Yatsimirsky A. Glabe C. Bioorg. Med. Chem. Let. 1999; 9: 2243-2248Crossref PubMed Scopus (143) Google Scholar, 28Phillips W.J. Cerione R.A. J. Biol. Chem. 1988; 263: 15498-15505Abstract Full Text PDF PubMed Google Scholar). Tyrosine fluorescence has been used to monitor the association of various metals with a number of proteins, including Aβ, α-transducin and, more recently, α-synuclein (27Garzon-Rodriguez W. Yatsimirsky A. Glabe C. Bioorg. Med. Chem. Let. 1999; 9: 2243-2248Crossref PubMed Scopus (143) Google Scholar, 28Phillips W.J. Cerione R.A. J. Biol. Chem. 1988; 263: 15498-15505Abstract Full Text PDF PubMed Google Scholar, 29Nielsen M.S. Vorum H. Lindersson E. Jensen P.H. J. Biol. Chem. 2001; 276: 22680-22684Abstract Full Text Full Text PDF PubMed Scopus (126) Google Scholar). In these studies tyrosine fluorescence is used as an indicator of changes in protein conformation or binding of metals. Exciting tyrosine at 280 nm elicits fluorescence that peaks at 310 nm for monomeric tyrosine and at 350–400 nm for tyrosinate (30Szabo A.G. Lynn K.R. Krajcarski D.T. Rayner D.M. FEBS Lett. 1978; 94: 249-252Crossref PubMed Scopus (72) Google Scholar). The fluorescence spectrum of α-synuclein yielded fluorescence peaks at 310 and 375 nm (Fig.1, A and B). Tyrosinate reactivity occurs when the phenolic hydroxyl group of tyrosine forms hydrogen bonds with carboxyl groups in nearby aspartates or glutamates. The fluorescence peak of α-synuclein at 375 nm showed a pH dependence similar to that of tyrosinate (Fig. 1C), which is consistent with the pH dependence of fluorescence due to tyrosinate. The peak at 375 nm had the highest intrinsic fluorescence at low pH and showed little change in fluorescence at pH > 7.0 (Fig. 1C). Further studies confirmed that the peak at 375 nm is not due to tyrosine dimerization, because both gel electrophoresis and mass spectrometry of the α-synuclein showed that the α-synuclein was monomeric (Fig. 1D, Coomassie gel of recombinant α-synuclein shown), and in addition, tyrosine dimerization of α-synuclein reduced its intrinsic fluorescence (Fig.1, E and F, described further below). These results suggest that the peak at 375 nm is due to tyrosinate, which could result from proton transfer from the phenolic hydroxyl to aspartic or glutamic acid protein acceptors (30Szabo A.G. Lynn K.R. Krajcarski D.T. Rayner D.M. FEBS Lett. 1978; 94: 249-252Crossref PubMed Scopus (72) Google Scholar). Next we used the fluorescence to examine the interaction of α-synuclein with metals. We observed three classes of interaction with α-synuclein. Class I metals included iron (Fe(II) and Fe(III)) and copper (Cu(II)) and decreased the fluorescence at both 310 and 375 nm (Fig. 1A). Class II metals included magnesium, zinc, and calcium, and increased the fluorescence at 375 nm, but did not affect the fluorescence at 310 nm (Figs.2A and 3A). Class III metals included nickel and manganese and did not affect α-synuclein fluorescence (data not shown). We proceeded to examine the fluorescence of α-synuclein in more detail to determine whether the metal induced changes α-synuclein fluorescence reflected interaction with metals or some other process, such as tyrosine dimerization. To examine whether the changes in fluorescence could be explained by tyrosine dimerization, we exposed monomeric α-synuclein to 312 nm light for 2 h (which is a process that induces tyrosine dimerization) and analyzed the emission fluorescence spectrum with excitation at either 280 or 315 nm. The emission spectrum derived using excitation at 280 nm showed that ultraviolet-irradiation reduced α-synuclein fluorescence strongly at 375 nm, but only weakly at 310 nm (Fig. 1E). The contrast between the changes in fluorescence induced by ultraviolet irradiation and by metals suggests that the changes in α-synuclein fluorescence induced by metals are not due to tyrosine dimerization. Analysis of the emission fluorescence spectrum of α-synuclein following excitation at 315 nm showed a reduced fluorescence at 380 nm for the ultraviolet-irradiated α-synuclein (Fig. 1F). In contrast, iron increased, rather than decreased, the fluorescence of α-synuclein as measured using the 315 nm excitation. These data indicate that the iron-induced quenching of α-synuclein fluorescence is not due to cross-linking of α-synuclein mediated by tyrosine dimerization.Figure 3The effects of zinc on α-synuclein fluorescence and dimerization. A, incubating zinc or calcium with recombinant α-synuclein increases the peak of α-synuclein emission fluorescence at 375 nm in a graded manner, using an excitation wavelength of 280 nm (λex = 280 nm; λem 290–450 nm). B, immunoblot of recombinant α-synuclein after incubation with zinc shows formation of a 32-kDa band consistent with formation of an SDS-resistant α-synuclein dimer. Lane 1, 0 nm ZnCl2; lane 2, 100 nmZnCl2; lane 3, 200 nmZnCl2; lane 4, 400 nmZnCl2.View Large Image Figure ViewerDownload Hi-res image Download (PPT) Plotting of the dose dependence of iron-induced fluorescence quenching showed a dose-dependent decrease in fluorescence, with an IC50 = 173 μm and a Hill coefficient of 1.0 (R2 = 1.0, p < 0.0001) (Fig. 1B), indicating one binding site or multiple binding sites with the same affinity and no cooperativity (Fig. 1, A and B). There is a small amount of binding of iron to α-synuclein between 1–10 μm Fe(II), which is a range that could be physiologically relevant (intracellular free iron is about 1.5 μm) (31Thomson A.M. Rogers J.T. Leedman P.J. Int. J. Biochem. Cell Biol. 1999; 31: 1139-1152Crossref PubMed Scopus (189) Google Scholar). The effect of magnesium on α-synuclein differed dramatically from that of iron. Magnesium increased the fluorescence at 375 nm but did not affect the fluorescence at 310 nm (Fig. 2A). Binding of magnesium to α-synuclein was also striking because the tyrosine fluorescence showed a sharp stepwise increase between 60 and 80 μm of magnesium indicating cooperative binding (Fig.2A). The cooperative regulation of tyrosine fluorescence, specifically at 375 nm, suggests that magnesium causes a conformational change in synuclein differing from that induced by iron. Co-incubating 80 μm magnesium with iron did not prevent iron-induced quenching of α-synuclein tyrosine fluorescence and in fact increased affinity of α-synuclein for iron from 173 to 50 μm(Fig. 2B). These data suggest that iron and magnesium bind to different sites on α-synuclein. Zinc and calcium also increased the fluorescence of α-synuclein at 375 nm, but showed a more graded pattern of interaction (Fig.3A). Jensen and colleagues recently noted a similar pattern of binding of calcium to α-synuclein (29Nielsen M.S. Vorum H. Lindersson E. Jensen P.H. J. Biol. Chem. 2001; 276: 22680-22684Abstract Full Text Full Text PDF PubMed Scopus (126) Google Scholar). Interestingly, immunoblots of recombinant α-synuclein following incubation with zinc showed that zinc induced formation of a prominent band at 32 kDa, consistent with formation of an SDS-resistant α-synuclein dimer (Fig. 3B). Neither magnesium nor calcium induced formation of SDS-resistant dimers identifiable by immunoblot (data not shown). We also examined how the A53T mutation in human α-synuclein affected binding of iron and magnesium. The A53T mutation did not change the apparent affinity of iron for α-synuclein (data not shown), but did abolish the interaction between magnesium and α-synuclein (Fig.2C). Previous studies have shown that the A53T mutation changes the conformation of α-synuclein by increasing its helical content (5Polymeropoulos M.H. Lavedan C. Leroy E. Ide S.E. Dehejia A. Dutra A. Pike B. Root H. Rubenstein J. Boyer R. Stenroos E.S. Chandrasekharappa S. Athanassiadou A. Papapetropoulos T. Johnson W.G. Lazzarini A.M. Di Duvoisin R.C. Iorio G. Golbe L.I. Nussbaum R.L. Science. 1997; 276: 2045-2047Crossref PubMed Scopus (6732) Google Scholar). These conformational changes might either reduce binding of magnesium to α-synuclein or prevent the conformational change associated with binding of magnesium to α-synuclein. The differing effects of magnesium and iron on the fluorescence spectrum of α-synuclein suggested to us that magnesium and iron might also induce different conformational states. We hypothesized that the conformational changes induced by binding of magnesium to α-synuclein might inhibit α-synuclein aggregation. To test this, we examined whether magnesium could inhibit the spontaneous aggregation of α-synuclein. α-Synuclein (23 μm) was incubated for 30 days at 37° ± MgCl2 (500 μm). To measure the amount of aggregation, the α-synuclein was diluted to 23 nm in the presence of 10 μm thioflavine-T (in 50 mm glycine pH 8.5), and the fluorescence spectrum was measured. The solution of aged α-synuclein showed a strong fluorescence peak at 480, indicating the presence of abundant β-pleated sheet structures (Fig.4A). Prior experiments have shown that the spontaneous aggregation of α-synuclein proceeds through a mechanism involving β-pleated sheet formation, and that thioflavine-T, which binds to proteins with β-pleated sheet structure, accurately measures α-synuclein aggregation (9Conway K.A. Lee S.J. Rochet J.C. Ding T.T. Williamson R.E. Lansbury Jr., P.T. Proc. Natl. Acad. Sci. U. S. A. 2000; 97: 571-576Crossref PubMed Scopus (1347) Google Scholar). Using thioflavine-T we observed that samples incubated in the presence of magnesium showed only base-line levels of fluorescence, indicating that magnesium prevented the formation of β-pleated sheet structures and the aggregation of α-synuclein (Fig. 4A). To verify that the magnesium was inhibiting α-synuclein aggregation, we measured the amount of aggregated α-synuclein in each sample by capturing the aggregates with 0.2-μm cellulose acetate filters, measuring the amount of retained α-synuclein by dot blot, and quantitating the resulting optical density. The results of the cellulose acetate assay paralleled the thioflavine-T assay and showed that magnesium prevented the spontaneous aggregation of α-synuclein (Fig. 4, B andC). Thus, two independent methods show that magnesium inhibits the spontaneous aggregation of α-synuclein in vitro. Magnesium was also able to prevent iron-induced α-synuclein aggregation. α-Synuclein (8 μm) was incubated with 50 μm FeCl2 for 72 h and then analyzed by thioflavine-T fluorescence or cellulose acetate. In both cases, α-synuclein samples co-incubated with 500 μm magnesium chloride showed little aggregation (Fig.5, A and B). The amount of thioflavine-T fluorescence induced by iron was less than that induced by spontaneously aggregated α-synuclein, which likely indicates that spontaneously induced α-synuclein contains more β-pleated sheet structure. Indeed analysis of iron-induced α-synuclein aggregates by circular dichroism did not show formation of β-pleated sheet structures, which suggests formation of a more amorphous aggregate (Fig. 5, C and D). These data suggest that magnesium inhibits the formation of α-synuclein aggregates containing either β-pleated sheet structure (via spontaneous aggregation) or amorphous structure (via iron-induced aggregation). We also examined aggregation by immunoblot analysis, which has been successfully used to examine aggregation of α-synuclein, as well as aggregation of other proteins implicated in neurodegenerative disease, such as the huntingtin and PrP proteins (29Nielsen M.S. Vorum H. Lindersson E. Jensen P.H. J. Biol. Chem. 2001; 276: 22680-22684Abstract Full Text Full Text PDF PubMed Scopus (126) Google Scholar, 32Giasson B. Uryu K. Trojanowski J. Lee V. J. Biol. Chem. 1999; 274: 7619-7622Abstract Full Text Full Text PDF PubMed Scopus (457) Google Scholar, 33Perrin R.J. Woods W.S. Clayton D.F. George J.M. J. Biol. Chem. 2000; 275: 34393-34398Abstract Full Text Full Text PDF PubMed Scopus (364) Google Scholar, 34Nucifora Jr., F.C. Sasaki M. Peters M.F. Huang H. Cooper J.K. Yamada M. Takahashi H. Tsuji S. Troncoso J. Dawson V.L. Dawson T.M. Ross C.A. Science. 2001; 291: 2423-2428Crossref PubMed Scopus (946) Google Scholar, 35Hegde R.S. Mastrianni J.A. Scott M.R. DeFea K.A. Tremblay P. Torchia M. DeArmond S.J. Prusiner S.B. Lingappa V.R. Science. 1998; 279: 827-834Crossref PubMed Scopus (617) Google Scholar). In these assays, wild-type recombinant α-synuclein (8 μm) was incubated with 0–3 mm FeCl2 and 0 or 100 μm MgCl2 for 24 h at 37 °C. The samples were immunoblotted with anti-α-synuclein antibody, and the total amount of α-synuclein reactivity above 46 kDa (which includes structures larger than a dimers) was quantified by video densitometry (Fig. 6, A and B). We observed a reduction in formation of high molecular weight immunoreactivity of α-synuclein at all but the highest dose of iron (n = 3, p < 0.0001). These results support the hypothesis that magnesium inhibits aggregation of α-synuclein induced following treatment with iron. Since we did not observe interaction between magnesium and A53T α-synuclein using tyrosine fluorescence, we tested whether aggregation of recombinant A53T α-synuclein was also insensitive to magnesium. We incubated recombinant A53T α-synuclein with 0–3 mm FeCl2 and 0 or 100 μmMgCl2 for 24 h, then immunoblotted the α-synuclein and quantified aggregation by video densitometry, as described above (Fig. 6, C and D). We did not observe consistent inhibition of iron-induced aggregation of A53T α-synuclein by magnesium. This suggests that magnesium cannot inhibit iron-induced aggregation of A53T α-synuclein. These data demonstrate that iron (II), magnesium, zinc, and calcium all interact with α-synuclein. Our data primarily rely on tyrosine fluorescence as a measure of the interaction of α-synuclein with metals. The Ki of α-synuclein for iron is 173 μm, and the Ki of magnesium for α-synuclein is between 60 and 80 μm. These affinities are consistent with an affinity of α-synuclein for calcium, determined by Nielson and colleagues (29Nielsen M.S. Vorum H. Lindersson E. Jensen P.H. J. Biol. Chem. 2001; 276: 22680-22684Abstract Full Text Full Text PDF PubMed Scopus (126) Google Scholar). Nielson and colleagues also confirmed their tyrosine fluorescence studies using equilibrium dialysis; our attempts at using equilibrium dialysis were stymied by extensive binding of α-synuclein to dialysis membranes. However, the studies of Nielson and colleagues show that tyrosine fluorescence provides an accurate indication of metal-synuclein binding interactions. The apparent affinity of α-synuclein for magnesium is strong enough to allow interaction of α-synuclein with magnesium in living cells, where the average intracellular concentration of magnesium is about 0.5 mm. This suggests that this interaction could have physiological significance. Although binding of magnesium to α-synuclein occurs at a concentration range that is physiologically significant, the concentration of free iron in the cell is much lower (<1.5 μm), which is far below the affinity of α-synuclein for iron that we observed (173 μm) (36Ponka P. Beaumont C. Richardson D. Semin. Hematol. 1998; 35: 35-54PubMed Google Scholar). However, studies using cell culture and neuropathology both suggest that α-synuclein interacts with iron. Incubating cells with iron induces α-synuclein aggregation in viable cells, which suggests that the concentration of iron in a cell is sufficient to induce α-synuclein aggregation under some conditions. In addition, α-synuclein aggregates in iron type I storage disease, and iron co-localizes with α-synuclein in Lewy bodies (19Galvin J.E. Giasson B. Hurtig H.I. Lee V.M. Trojanowski J.Q. Am. J. Pathol. 2000; 157: 361-368Abstract Full Text Full Text PDF PubMed Scopus (170) Google Scholar). Although it is unclear how α-synuclein might interact with iron in the living cell, it is possible that cofactors increase the affinity of α-synuclein for iron sufficient to allow a physiological interaction. Many other factors might also affect the behavior of α-synuclein. For instance, binding to lipids and phosphorylation or binding to β-synuclein have all been shown to change the biochemistry of α-synuclein, and these agents might increase its affinity for iron (44Hashimoto M. Rockenstein E. Mante M. Mallory M. Masliah E. Neuron. 2001; 32: 213-223Abstract Full Text Full Text PDF PubMed Scopus (370) Google Scholar). Our preliminary studies examining magnesium already provide a hint of modulation. TheKi of iron (II) drops to 50 μm in the presence of magnesium. Future studies might unravel the biochemistry of α-synuclein further. Although binding of magnesium appears to introduce a conformation that promotes binding of iron, this same conformational change inhibits aggregation of α-synuclein. We hypothesize that magnesium either changes the conformation of α-synuclein to one that resists aggregation or induces dimerization to a structure that resists aggregation (Fig. 7). The ability of zinc to induce SDS-resistant α-synuclein dimers, coupled with the similarity the changes in tyrosine fluorescence observed with magnesium and zinc, suggest that magnesium might induce dimerization of α-synuclein in a manner similar to that of zinc. Future studies using nuclear magnetic resonance spectroscopy will need to be performed to investigate further how magnesium affects the conformation of α-synuclein. The most important observation made in this paper is that magnesium inhibits the aggregation of α-synuclein. This observation is supported by our use of four independent lines of investigation (immunoblot, cellulose acetate filtration, and thioflavine-T fluorescence). The type of aggregate measured by each assay likely differs slightly. Immunoblotting detects aggregates that are stable enough to resist both heating and SDS. Cellulose acetate filtration and thioflavine-T are more gentle methods that can detect both stable aggregates and also aggregates that might be re-dissolved by SDS. Thioflavine-T recognizes aggregate with a β-pleated sheet structure. Interestingly, spontaneously aggregated α-synuclein shows much more fluorescence by thioflavine-T than iron-induced aggregate, suggesting that the former has more β-pleated sheet structure. We have also taken care to examine two forms of α-synuclein aggregation: spontaneous and iron-induced aggregation. Many studies show that α-synuclein has a strong tendency to spontaneously aggregate, and this is the most widely accepted method for inducing α-synuclein aggregation (7Hashimoto M. Hsu L. Sisk A. Xia Y. Takeda A. Sundsmo M. Masliah E. Brain Res. 1998; 799: 301-306Crossref PubMed Scopus (249) Google Scholar, 8Wood S. Wypych J. Steavenson S. Louis J. Citron M. Biere A. J. Biol. Chem. 1999; 274: 19509-19512Abstract Full Text Full Text PDF PubMed Scopus (612) Google Scholar, 9Conway K.A. Lee S.J. Rochet J.C. Ding T.T. Williamson R.E. Lansbury Jr., P.T. Proc. Natl. Acad. Sci. U. S. A. 2000; 97: 571-576Crossref PubMed Scopus (1347) Google Scholar). Metal-induced aggregation has only been investigated by a small number of groups, but is perhaps the only method currently available for inducing α-synuclein aggregation in cultured cells (10Ostrerova-Golts N. Petrucelli L. Hardy J. Lee J. Farrer M. Wolozin B. J. Neurosci. 2000; 20: 6048-6054Crossref PubMed Google Scholar, 15Paik S.R. Shin H.J. Lee J.H. Arch. Biochem. Biophys. 2000; 378: 269-277Crossref PubMed Scopus (173) Google Scholar, 37LJ M. Hashimoto H. Xia Y. Takeda A. Sisk A. Sundsmo M. Masliah E. NeuroReport. 1999; 10: 717-721Crossref PubMed Scopus (385) Google Scholar). The ability of magnesium to inhibit α-synuclein aggregation induced by both protocols (spontaneous and iron-induced) suggests that this is a robust phenomenon. Increasing evidence suggests that metals play a pivotal role in the pathophysiology of neurodegenerative disorders. Zinc and copper greatly accelerate aggregation of β-amyloid and might play a critical role in neurotoxicity induced by β-amyloid (38Atwood C. Moir R. Huang X. Scarpa R. Bacarra N. Romano D. Hartshorn M. Tanzi R. Bush A. J. Biol. Chem. 1998; 273: 12817-12826Abstract Full Text Full Text PDF PubMed Scopus (945) Google Scholar, 39Bush A.I. Pettingell W.H. Multhaup G. P. M.d P.M. Vonsattel J.P. Gusella J.F. Beyreuther K. Masters C.L. Tanzi R.E. Science. 1994; 265: 1464-1467Crossref PubMed Scopus (1411) Google Scholar). Copper and manganese both bind to the prion protein and appear to influence the clinical course of prion-induced neurodegeneration (40Brown D.R. Qin K. Herms J.W. Madlung A. Manson J. Strome R. Fraser P.E. Kruck T. von Bohlen A. Schulz-Schaeffer W. Giese A. Westaway D. Kretzschmar H. Nature. 1997; 390: 684-687Crossref PubMed Scopus (37) Google Scholar, 41Wadsworth J.D. Hill A.F. Joiner S. Jackson G.S. Clarke A.R. Collinge J. Nat. Cell Biol. 1999; 1: 55-59Crossref PubMed Scopus (280) Google Scholar). Iron levels are increased in brains of patients with PD, and iron is present in Lewy bodies. Neuromelanin selectively binds Fe(III) and might liberate Fe(II) as the Fe(III) is reduced to Fe(II) (via the Haber Weiss reaction) by free radicals produced in response to the oxidative stress associated with PD, providing a potential source of Fe(II) to accelerate α-synuclein aggregation (42Hirsch E.C. Graybiel A.M. Agid Y. Acta Neurol. Scand. Suppl. 1989; 126: 19-22Crossref PubMed Scopus (62) Google Scholar, 43Wolozin B. Golts N. Neuroscientist. 2001; 8: 22-32Crossref Scopus (80) Google Scholar). On the other hand, magnesium levels are reduced in brains of patients with PD (21Mann V.M. Cooper J.M. Daniel S.E. Srai K. Jenner P. Marsden C.D. Schapira A.H. Ann. Neurol. 1994; 36: 876-881Crossref PubMed Scopus (216) Google Scholar, 22Griffiths P.D. Dobson B.R. Jones G.R. Clarke D.T. Brain. 1999; 122: 667-673Crossref PubMed Scopus (198) Google Scholar, 23Dexter D. Carayon A. Javoy-Agid F. Agid Y. Wells F. Daniel S. Lees A. Jenner P. Marsden C. Brain. 1991; 114: 1953-1975Crossref PubMed Scopus (871) Google Scholar, 24Barbiroli B. Martinelli P. Patuelli A. Lodi R. Iotti S. Cortelli P. Montagna P. Mov. Disord. 1999; 14: 430-435Crossref PubMed Scopus (51) Google Scholar, 25Uitti R.J. Rajput A.H. Rozdilsky B. Bickis M. Wollin T. Yuen W.K. Can. J. Neurol. Sci. 1989; 16: 310-314Crossref PubMed Scopus (111) Google Scholar, 26Durlach J. Bac P. Durlach V. Rayssiguier Y. Bara M. Guiet-Bara A. Magnes. Res. 1998; 11: 25-42PubMed Google Scholar). If iron accelerates α-synuclein aggregation, then the abundance of iron in the substantia nigra could increase the tendency of Lewy bodies to accumulate in this region. In a companion paper, 2N. Golts, H. Snyder, M. Frasier, C. Theisler, P. Choi, and B. Wolozin, submitted for publication. we extend our studies from the test tube to neurons, and show that magnesium also inhibits aggregation of α-synuclein in neurons. Together, these data raise the possibility that iron and magnesium might also modulate α-synuclein aggregation in the brain. We acknowledge the assistance of Alan Frankfater (Loyola University) and Brian Shoichet (Northwestern) and thank Matthew Farrer and John Hardy (Mayo Clinic) for providing the α-synuclein cDNA.
Referência(s)